Search

We develop the theory of the non-adiabatic phonon self-energy arising from coupling to electrons with finite linewidths using the spectral representation of Green's functions. Our formalism naturally includes the contribution from the intra-band electronic transitions at all electron wavevectors to the phonon linewidths, which is forbidden for zone-center optical phonons if infinitesimal electron linewidths are used. As a proof of principle, we use the theory to calculate the linewidth of the double-degenerate, zone-center optical $E_{\rm 2g}$ phonons of graphene as a function of the chemical potential. The calculated phonon linewidths obtained with finite electron linewidths differ significantly from those obtained with infinitesimal electron linewidths even at low temperatures. Intra-band electronic transitions play an important role in making this difference. Moreover, only the results accounting for the finite electron linewidths are in quantitative agreement with available experimental data. The presented formalism is suitable for first-principles calculations., Comment: 11 pages, 3 figures; any comments, especially on missed works, are welcome

Verification and validation of methods and first-principles software are at the core of computational solid-state physics but are too rarely addressed. We compare four first-principles codes: Abinit, Quantum ESPRESSO, EPW, ZG, and three methods: (i) the Allen-Heine-Cardona theory using density functional perturbation theory (DFPT), (ii) the Allen-Heine-Cardona theory using Wannier function perturbation theory (WFPT), and (iii) an adiabatic non-perturbative frozen-phonon method. For these cases, we compute the real and imaginary parts of the electron-phonon self-energy in diamond and BAs, including dipoles and quadrupoles when interpolating. We find excellent agreement between software that implements the same formalism as well as good agreement between the DFPT and WFPT methods. Importantly, we find that the Deybe-Waller term is momentum dependent which impacts the mass enhancement, yielding approximate results when using the Luttinger approximations. Finally, we compare the electron-phonon spectral functions between Abinit and EPW and find excellent agreement even away from the band edges., Comment: 20 pages and 10 figures

Acoustic phonons in piezoelectric materials strongly couple to electrons through a macroscopic electric field. We show that this coupling leads to a momentum-dependent divergence of the Fan-Migdal electron linewidth. We then develop a self-consistent theory for calculating electron linewidths, which not only removes this piezoelectric divergence but also considerably modifies the linewidth in nonpiezoelectric, polar materials. Our predictions await immediate experimental confirmation, and this self-consistent method should be broadly used in interpreting various experiments on the electronic properties of real materials.

Although plasmons and phonons are the collective excitations that govern the low-energy physics of doped semiconductors, their nonadiabatic hybridization and mutual screening have not been studied from first principles. We achieve this goal by transforming the Dyson equation to the frequency-independent dynamical matrix of an equivalent damped oscillator. Calculations on doped GaAs and TiO2 agree well with available Raman data and await immediate experimental confirmation from infrared, neutron, electron-energy-loss, and angle-resolved photoemission spectroscopies.

The nonlinear Hall effect has attracted much attention due to the famous, widely adopted interpretation in terms of the Berry curvature dipole in momentum space. Using ab initio Boltzmann transport equations, we find a 60% enhancement in the nonlinear Hall effect of p-doped GeTe and its noticeable frequency dependence, qualitatively different from the predictions based on the Berry curvature dipole. The origin of these differences is long-lived valley polarization in the electron distribution arising from electron-phonon scattering. Our findings await immediate experimental confirmation.

Many physical quantities in solid-state physics are calculated from k-space summation. For spectral functions, the frequency-dependent factor can be decomposed into the energy-conserving delta function part and the nondissipative principal value part. A very useful scheme for this k-space summation is the tetrahedron method. Tetrahedron methods have been widely used to calculate the summation of the energy-conserving delta function part such as the imaginary part of the dielectric function. On the other hand, the corresponding tetrahedron method for the nondissipative part such as the real part of the dielectric function has not been used much. In this paper, we address the technical difficulties in the tetrahedron method for the nondissipative part and present an easy-to-implement, stable method to overcome those difficulties. We demonstrate our method by calculating the static and dynamical spin Hall conductivity of platinum. Our method can be widely applied to calculate linear static or dynamical conductivity, self-energy of an electron, and electric polarizability, to name a few., Comment: 11 pages, 6 figures, 1 table

The collapse of Landau levels under an electric field perpendicular to the magnetic field is one of the distinctive features of Dirac materials. So is the coupling of lattice deformations to the electronic degrees of freedom in the form of gauge fields which allows the formation of pseudo-Landau levels from strain. We analyze the collapse of Landau levels induced by strain on realistic Weyl semimetals hosting anisotropic, tilted Weyl cones in momentum space. We perform first-principles calculations, to establish the conditions on the external strain for the collapse of Landau levels in TaAs which can be experimentally accessed., Comment: 9 pages, 6 figures

In recent years, generative adversarial network (GAN)-based image generation techniques design their generators by stacking up multiple residual blocks. The residual block generally contains a shortcut, \ie skip connection, which effectively supports information propagation in the network. In this paper, we propose a novel shortcut method, called the gated shortcut, which not only embraces the strength point of the residual block but also further boosts the GAN performance. More specifically, based on the gating mechanism, the proposed method leads the residual block to keep (or remove) information that is relevant (or irrelevant) to the image being generated. To demonstrate that the proposed method brings significant improvements in the GAN performance, this paper provides extensive experimental results on the various standard datasets such as CIFAR-10, CIFAR-100, LSUN, and tiny-ImageNet. Quantitative evaluations show that the gated shortcut achieves the impressive GAN performance in terms of Frechet inception distance (FID) and Inception score (IS). For instance, the proposed method improves the FID and IS scores on the tiny-ImageNet dataset from 35.13 to 27.90 and 20.23 to 23.42, respectively., Comment: arXiv admin note: text overlap with arXiv:2112.14968

The spin photocurrents, direct currents induced by light, hold great promise for introducing new elements to spintronics. However, a general theory for spin photocurrents in real materials which is applicable to systems with spin-orbit coupling or noncollinear magnetism is absent. Here, we develop such a general theory of second-order spin photocurrents. We find that the secondorder spin photocurrents can be classified into Drude, Berry curvature dipole, shift, injection, and rectification currents, which have different physical origins and symmetry properties. Surprisingly, our theory predicts a direct pure spin rectification current in an insulator induced by photons with energies lower than the material band gap. This phenomenon is absent in the case of the charge photocurrent. We find that the pure spin current of BiTeI induced by subgap light is large enough to be observable in experiments. Moreover, the subgap pure spin photocurrent is highly tunable with the polarization of light and the flowing direction of the spin photocurrent. This study lays the groundwork for the study of nonlinear spin photocurrents in real materials and provides a route to engineer light-controlled spin currents., Comment: 25 pages, 11 figures

Thanks to the nearsightedness principle, the low-energy electronic structure of solids can be represented by localized states such as the Wannier functions. Wannier functions are actively being applied to a wide range of phenomena in condensed matter systems. However, the Wannier-functionbased representation is limited to a small number of bands and thus cannot describe the change of wavefunctions due to various kinds of perturbations, which require sums over an infinite number of bands. Here, we introduce the concept of the Wannier function perturbation, which provides a localized representation of wavefunction perturbations. Wannier function perturbation theory allows efficient calculation of numerous quantities involving wavefunction perturbation, among which we provide three applications. First, we calculate the temperature-dependent indirect optical absorption spectra of silicon near the absorption edge nonadiabatically, i.e., differentiating phonon-absorption and phonon-emission processes, and without arbitrary temperature-dependent shifts in energy. Second, we establish a theory to calculate the shift spin conductivity without any band-truncation error. Unlike the shift charge conductivity, an exact calculation of the shift spin conductivity is not possible within the conventional Wannier function methods because it cannot be obtained from geometric quantities for low-energy bands. We apply the theory to monolayer WTe$_2$. Third, we calculate the spin Hall conductivity of the same material again without any band-truncation error. Wannier function perturbation theory is a versatile method that can be readily applied to calculate a wide range of quantities related to various kinds of perturbations., Comment: 31+6 pages, 14+7 figures

Here, a magnetic model with an unprecedentedly large number of parameters was determined from first-principles calculations for transition-metal phosphorus trisulfides (TMPS$_3$'s), which reproduced the measured magnetic ground states of bulk TMPS$_3$'s. Our Monte Carlo simulations for the critical temperature, magnetic susceptibility, and specific heat of bulk and few-layer TMPS$_3$'s agree well with available experimental data and show that the antiferromagnetic order of FePS$_3$ and NiPS$_3$ persists down to monolayers. Remarkably, the orbital polarization, which was neglected in recent first-principles studies, dramatically enhances the magnetic anisotropy of FePS$_3$ by almost two orders of magnitude. A recent Raman study [K. Kim et al., Nat. Commun. 10, 345 (2019)] claimed that magnetic ordering is absent in monolayer NiPS$_3$ but simultaneously reported a strong two-magnon continuum; we show that the criterion used to judge magnetic ordering there is invalid in monolayer NiPS$_3$, thus providing an understanding of the two seemingly contradictory experimental results. The rich predictions on the magnetic susceptibility and specific heat of few-layer FePS$_3$ and NiPS$_3$ await immediate experimental verifications., Comment: This document is the unedited Author's version of a Submitted Work that was subsequently accepted for publication in Nano Letters, American Chemical Society after peer review. To access the final edited and published work see https://pubs.acs.org/articlesonrequest/AOR-3TCU3CMMVGHMTTRIVWYS

This paper presents the predictive accuracy using two-variate meteorological factors, average temperature and average humidity, in neural network algorithms. We analyze result in five learning architectures such as the traditional artificial neural network, deep neural network, and extreme learning machine, long short-term memory, and long-short-term memory with peephole connections, after manipulating the computer-simulation. Our neural network modes are trained on the daily time-series dataset during seven years (from 2014 to 2020). From the trained results for 2500, 5000, and 7500 epochs, we obtain the predicted accuracies of the meteorological factors produced from outputs in ten metropolitan cities (Seoul, Daejeon, Daegu, Busan, Incheon, Gwangju, Pohang, Mokpo, Tongyeong, and Jeonju). The error statistics is found from the result of outputs, and we compare these values to each other after the manipulation of five neural networks. As using the long-short-term memory model in testing 1 (the average temperature predicted from the input layer with six input nodes), Tonyeong has the lowest root mean squared error (RMSE) value of 0.866 $(%)$ in summer from the computer-simulation in order to predict the temperature. To predict the humidity, the RMSE is shown the lowest value of 5.732 $(%)$, when using the long short-term memory model in summer in Mokpo in testing 2 (the average humidity predicted from the input layer with six input nodes). Particularly, the long short-term memory model is is found to be more accurate in forecasting daily levels than other neural network models in temperature and humidity forecastings. Our result may provide a computer-simuation basis for the necessity of exploring and develping a novel neural network evaluation method in the future., Comment: 22 pages, 8 figures, 4 Tables. arXiv admin note: substantial text overlap with arXiv:2101.09356

It is important to calculate and analyze temperature and humidity prediction accuracies among quantitative meteorological forecasting. This study manipulates the extant neural network methods to foster the predictive accuracy. To achieve such tasks, we analyze and explore the predictive accuracy and performance in the neural networks using two combined meteorological factors (temperature and humidity). Simulated studies are performed by applying the artificial neural network (ANN), deep neural network (DNN), extreme learning machine (ELM), long short-term memory (LSTM), and long short-term memory with peephole connections (LSTM-PC) machine learning methods, and the accurate prediction value are compared to that obtained from each other methods. Data are extracted from low frequency time-series of ten metropolitan cities of South Korea from March 2014 to February 2020 to validate our observations. To test the robustness of methods, the error of LSTM is found to outperform that of the other four methods in predictive accuracy. Particularly, as testing results, the temperature prediction of LSTM in summer in Tongyeong has a root mean squared error (RMSE) value of 0.866 lower than that of other neural network methods, while the mean absolute percentage error (MAPE) value of LSTM for humidity prediction is 5.525 in summer in Mokpo, significantly better than other metropolitan cities., Comment: 33 Pages, 9 figures

With the advanced investigations into low-dimensional systems, it has become essential to find materials having interesting lattices that can be exfoliated down to monolayer. One particular important structure is a kagome lattice with its potentially diverse and vibrant physics. We report a van-der-Waals kagome lattice material, Pd3P2S8, with several unique properties such as an intriguing flat band. The flat band is shown to arise from a possible compact-localized state of all five 4d orbitals of Pd. The diamagnetic susceptibility is precisely measured to support the calculated susceptibility obtained from the band structure. We further demonstrate that Pd3P2S8 can be exfoliated down to monolayer, which ultimately will allow the possible control of the localized states in this two-dimensional kagome lattice using the electric field gating.

The anharmonic lattice is a representative example of an interacting bosonic many-body system. The self-consistent harmonic approximation has proven versatile for the study of the equilibrium properties of anharmonic lattices. However, the study of dynamical properties therewithin resorts to an ansatz, whose validity has not yet been theoretically proven. Here, we apply the time-dependent variational principle, a recently emerging useful tool for studying the dynamic properties of interacting many-body systems, to the anharmonic lattice Hamiltonian at finite temperature using the Gaussian states as the variational manifold. We derive an analytic formula for the position-position correlation function and the phonon self-energy, proving the dynamical ansatz of the self-consistent harmonic approximation. We establish a fruitful connection between time-dependent variational principle and the anharmonic lattice Hamiltonian, providing insights in both fields. Our work expands the range of applicability of time-dependent variational principle to first-principles lattice Hamiltonians and lays the groundwork for the study of dynamical properties of the anharmonic lattice using a fully variational framework., Comment: v2: Added a citation to L. Monacelli and F. Mauri, "Time-Dependent Self Consistent Harmonic Approximation: Anharmonic nuclear quantum dynamics and time correlation functions," arXiv:2011.14986 and a note on it

Quasiparticle interference (QPI) patterns in momentum space are often assumed to be independent of the strength of the impurity potential when compared with other quantities, such as the joint density of states. Here, using the $T$-matrix theory, we show that this assumption breaks down completely even in the simplest case of a single-site impurity on the square lattice with an $s$ orbital per site. Then, we predict from first-principles, a very rich, impurity-strength-dependent structure in the QPI pattern of TaAs, an archetype Weyl semimetal. This study thus demonstrates that the consideration of the details of the scattering impurity including the impurity strength is essential for interpreting Fourier-transform scanning tunneling spectroscopy experiments in general., Comment: main manuscript: 8 pages, 6 figures, Supplementary Information: 3 pages, 6 figures

The Allen-Heine-Cardona theory allows us to calculate phonon-induced electron self-energies from first principles without resorting to the adiabatic approximation. However, this theory has not been able to account for the change of the electron wave function, which is crucial if interband energy differences are comparable to the phonon-induced electron self-energy as in temperature-driven topological transitions. Furthermore, for materials without inversion symmetry, even the existence of such topological transitions cannot be investigated using the Allen-Heine-Cardona theory. Here, we generalize this theory to the renormalization of both the electron energies and wave functions. Our theory can describe both the diagonal and off-diagonal components of the Debye-Waller self-energy in a simple, unified framework. For demonstration, we calculate the electron-phonon coupling contribution to the temperature-dependent band structure and hidden spin polarization of BiTlSe2 across a topological transition. These quantities can be directly measured. Our theory opens a door for studying temperature-induced topological phase transitions in materials both with and without inversion symmetry., Comment: v3: Typo fixed in Eq.(5)

3D hand pose estimation from a single depth image plays an important role in computer vision and human-computer interaction. Although recent hand pose estimation methods using convolution neural network (CNN) have shown notable improvements in accuracy, most of them have a limitation that they rely on a complex network structure without fully exploiting the articulated structure of the hand. A hand, which is an articulated object, is composed of six local parts: the palm and five independent fingers. Each finger consists of sequential-joints that provide constrained motion, referred to as a kinematic chain. In this paper, we propose a hierarchically-structured convolutional recurrent neural network (HCRNN) with six branches that estimate the 3D position of the palm and five fingers independently. The palm position is predicted via fully-connected layers. Each sequential-joint, i.e. finger position, is obtained using a recurrent neural network (RNN) to capture the spatial dependencies between adjacent joints. Then the output features of the palm and finger branches are concatenated to estimate the global hand position. HCRNN directly takes the depth map as an input without a time-consuming data conversion, such as 3D voxels and point clouds. Experimental results on public datasets demonstrate that the proposed HCRNN not only outperforms most 2D CNN-based methods using the depth image as their inputs but also achieves competitive results with state-of-the-art 3D CNN-based methods with a highly efficient running speed of 285 fps on a single GPU.

We present a method to compute the intrinsic spin Hall conductivity from first principles using an interpolation scheme based on maximally-localized Wannier functions. After obtaining the relevant matrix elements among the ab initio Bloch states calculated on a coarse k-point mesh, we Fourier transform them to find the corresponding matrix elements between Wannier states. We then perform an inverse Fourier transform to interpolate the velocity and spin-current matrix elements onto a dense k-point mesh, and use them to evaluate the spin Hall conductivity as a Brillouin-zone integral. This strategy has a much lower computational cost than a direct ab initio calculation, without sacrificing the accuracy. We demonstrate that the spin Hall conductivities of platinum and doped gallium arsenide, computed with our interpolation scheme as a function of the Fermi energy, are in good agreement with those obtained in previous first-principles studies. We also discuss certain approximations that can be made, in the spirit of the tight-binding method, to simplify the calculation of the velocity and spin-current matrix elements in the Wannier representation., Comment: 13 pages, 6 figures

Magnetic ordering in the two-dimensional limit has been one of the most important issues in condensed matter physics for the past several decades. The recent discovery of new magnetic van der Waals materials heralds a much-needed easy route for the studies of two-dimensional magnetism: the thickness dependence of the magnetic ordering has been examined by using Ising- and XXZ-type magnetic van der Waals materials. Here, we investigated the magnetic ordering of MnPS3, a two-dimensional antiferromagnetic material of Heisenberg-type, by Raman spectroscopy from bulk all the way down to bilayer. The phonon modes that involve the vibrations of Mn ions exhibit characteristic changes as the temperature gets lowered through the N\'eel temperature. In bulk MnPS3, the Raman peak at ~155 cm-1 becomes considerably broadened near the N\'eel temperature and upon further cooling is subsequently red-shifted. The measured peak positions and polarization dependences of the Raman spectra are in excellent agreement with our first-principles calculations. In few-layer MnPS3, the peak at ~155 cm-1 exhibits the characteristic red-shift at low temperatures down to the bilayer, indicating that the magnetic ordering is surprisingly stable at such a thin limit. Our work sheds light on the hitherto unexplored magnetic ordering in the Heisenberg-type antiferromagnetic systems in the atomic-layer limit., Comment: 22 pages, 3 figures, supplementary information 21 pages

How a certain ground state of complex physical systems emerges, especially in two-dimensional materials, is a fundamental question in condensed-matter physics. A particularly interesting case is systems belonging to the class of XY Hamiltonian where the magnetic order parameter of conventional nature is unstable in two-dimensional materials leading to a Berezinskii-Kosterlitz-Thouless transition. Here, we report how the XXZ-type antiferromagnetic order of a magnetic van der Waals material, NiPS3, behaves upon reducing the thickness and ultimately becomes unstable in the monolayer limit. Our experimental data are consistent with the findings based on renormalization group theory that at low temperatures a two-dimensional XXZ system behaves like a two-dimensional XY one, which cannot have a long-range order at finite temperatures. This work provides experimental examination of the XY magnetism in the atomically thin limit and opens new opportunities of exploiting these fundamental theorems of magnetism using magnetic van der Waals materials., Comment: 57 pages, 24 figures (including Supplementary Information)

Maximally localized Wannier functions are localized orthogonal functions that can accurately represent given Bloch eigenstates of a periodic system at a low computational cost, thanks to the small size of each orbital. Tight-binding models based on the maximally localized Wannier functions obtained from different systems are often combined to construct tight-binding models for large systems such as a semi-infinite surface. However, the corresponding maximally localized Wannier functions in the overlapping region of different systems are not identical, and this discrepancy can introduce serious artifacts to the combined tight-binding model. Here, we propose two methods to seamlessly stitch two different tight-binding models that share some basis functions in common. First, we introduce a simple and efficient method: (i) finding the best matching maximally localized Wannier function pairs in the overlapping region belonging to the two tight-binding models, (ii) rotating the spin orientations of the two corresponding Wannier functions to make them parallel to each other, and (iii) making their overall phases equal. Second, we propose a more accurate and generally applicable method based on the iterative minimization of the difference between the Hamiltonian matrix elements in the overlapping region. We demonstrate our methods by applying them to the surfaces of diamond, GeTe, Bi$_2$Se$_3$, and TaAs. Our methods can be readily used to construct reliable tight-binding models for surfaces, interfaces, and defects., Comment: 18 pages, 11 figures